Understanding the Role of Artificial Intelligence in Personalized Engagement Marketing

Part of: Digital Marketing Topic 10 — Optimization, Data & Measurement · Reading Citation: Kumar, V., Rajan, B., Venkatesan, R., & Lecinski, J. (2019). Understanding the role of artificial intelligence in personalized engagement marketing. California Management Review, 61(4), 135–155. Key concepts: Personalization, Curation, Customer Engagement, Artificial Intelligence, Uncanny Valley

---

TL;DR

Kumar et al. explore how AI enables personalized engagement marketing — creating, communicating and delivering personalised offerings to customers. Their core claim: consumers have entered a "Wave 3" journey in which AI curates endless options into personalised, relevant choices. The article presents an integrative framework, distinguishes personalization (firm-controlled) from customization (customer-controlled), and offers managerial predictions across two dimensions — time (short vs long run) and place (developed vs developing economies) — built around the three CRM pillars of acquisition, retention and growth.

Why It's on the Reading List

It is the data/personalisation anchor for the optimization topic: it explains how AI-driven curation reshapes the customer journey and CRM strategy, introduces the "three waves" model and the developed/developing economy predictions, and supplies the readiness checklist plus the uncanny-valley caveat. It is also the marketing-discipline definition of AI cited across the AI-advertising readings.

Background & Research Question

Technological shifts (PCs, Internet, smartphones) restructure firm strategy. AI is the latest, delivering value mainly through personalisation. The article asks: given AI algorithms and increasingly cognitive consumer interfaces, how should brand managers approach personalization — and what does AI mean for two key strategic assets (brands and customers) across developed and developing economies?

Definition — Artificial intelligence (Kaplan & Haenlein, used here)

"A system's ability to interpret external data correctly, to learn from such data, and to use those learnings to achieve specific goals and tasks through flexible adaptation." AI operates in the domain of automation and continuous learning.

Key Concepts & Definitions

Personalization vs Customization

Personalization = the firm decides the suitable marketing mix for the individual, based on previously collected customer data (firm-controlled). Customization = the customer proactively specifies one or more marketing-mix elements (customer-decided).

Definition — Curation (in AI-driven engagement)

The automatic, machine-driven selection of products, prices, website content and advertising messages that fit an individual customer's preferences. AI predicts the type, timing and purchase of preferred offerings, reducing consumer cognitive load.

Definition — Customer Engagement (CE)

The attitude, behaviour and level of connectedness among customers, and between customers and employees/firm. Comprises four dimensions: CLV (lifetime value), CRV (referral value), CIV (influence value), CKV (knowledge value).

Framework / Model

Three waves of customer information processing:

Why Wave 3 emerged

Wave 2 created the paradox of choice and digital cognitive load — too much information from configurators/search. AI resolves this by curating, addressing both the First Moment of Truth (shelf) and the Zero Moment of Truth (online).

The integrative framework links consumer choice/decision-making and knowledge organisation to AI-driven curation of the marketing mix (product, price, place, promotion), moderated by AI-driven capabilities — marketing (customer-level data), technological (system compatibility), and operational (continuous sensing/learning).

Main Arguments / Findings

What AI offers firms (four benefits):

1. Revenue growth (not just cost-cutting) — e.g. Toyota's $4bn AI/robotics institute; Baidu's $1.9bn AI finance arm.
2. Product curation at superhuman scale — real-time, accurate, non-disruptive (IBM scalable deep learning; eBay predictive matching).
3. Frees managers for creativity (L'Oréal social listening; McCormick + IBM Watson for spice R&D).
4. Broad applicability — McKinsey estimates AI value of $1.4–2.6tn in marketing & sales and $1.2–2.0tn in supply chain/manufacturing.

Is AI a magic pill? — Readiness conditions

AI is not automatic value. Firms need: (1) data maturity (well-connected data ecosystem + data scientists); (2) alignment with firm goals (organisation-wide, possibly interdisciplinary not top-down); (3) clear control parameters (the Turing-test problem of interpretability); (4) workforce transformation readiness; (5) handling of ethical and privacy concerns (what data can be used, expunged; human bias fed into AI → confirmation bias).

Predictions: time × place

Short run — tactical solutions.

• Developed economies: superior brand experiences (Spotify Discover Weekly), dynamic pricing (Jet), automated service (Uber Eats), content-focused ads (McCann AI creative director).
• Developing economies: brand trust (Juntos financial messaging in 15 countries), affordable pricing (Amazon India), standardised service (Keeko teaching robot, China), solutions-focused ads (Ogilvy/Nestlé nutrition assistant).

Long run — developed economies (the three CRM pillars):

Long run — developing economies: regional brand value (local culture/language, e.g. Google's Neighbourly app for India); incremental & adaptive human–machine interface (job-displacement anxiety, slow macro adoption); profitable customer loyalty (sparse CLV data; loyalty programmes rewarding higher spend).

Definition — Uncanny valley

The hypothesised dip in observers' affinity when a humanoid object appears almost but not exactly human, eliciting eeriness/revulsion. Managers must balance making AI assistants human-like vs retaining some artificiality. Related: algorithm aversion — people trust human forecasters over machines even when machines are more accurate.

Implications for Marketers

• In an AI/curation world, brand value becomes more important, not less — consumers must seek out the brand directly when intermediaries (Amazon, Google, voice) control discovery.
• Collect first-party consumption data to feed recommendation algorithms (direct apps, chatbots, connected products).
• Rethink the customer portfolio: knowledge value (CKV) customers — those who most improve algorithm accuracy — need not be high-CLV/CRV customers.
• Tailor strategy by economy: hyper-local/regional branding and incremental AI in developing markets; sophisticated multisensory interfaces (mindful of the uncanny valley) in developed markets.

Exam Takeaways

Likely exam points

• Personalization (firm-controlled) vs Customization (customer-controlled).
• The three waves and why Wave 3 emerged (paradox of choice, digital cognitive load).
• Curation as machine-driven selection of the marketing mix.
• The four CE dimensions: CLV, CRV, CIV, CKV — and why CKV matters in an AI world.
• The two prediction dimensions: time (short/long run) × place (developed/developing) and the three CRM pillars (acquisition/retention/growth).
• Uncanny valley and algorithm aversion as retention challenges.
• "Is AI a magic pill?" — the five readiness conditions (data maturity, goal alignment, control, workforce, ethics/privacy).

Summary

• Conceptual/managerial article (CMR), with an integrative framework + many illustrative cases.
• AI delivers value chiefly through personalised curation, moving consumers into "Wave 3".
• AI is not a magic pill — requires data maturity, alignment, control, workforce change and ethics.
• Long-run impact mapped across developed/developing economies via acquisition, retention and growth.

Related Notes

• Digital Marketing — subject hub
• Personalization · Curation — the twin mechanisms
• Customer Engagement — the CLV/CRV/CIV/CKV construct
• Uncanny Valley — the human–machine interface caveat
• 09a-Rodgers - Promises and Perils of AI and Advertising — cites this paper as the marketing definition of AI